Recommendations

SFED ConsensusAcute lower gastrointestinal haemorrhage

Head of project: JP Arpurt, G Lesur

Task committee: D Heresbach, D Soudan, P Burtin

Reading committee: T Barrioz, A Barkun, R Laugier and JP Arpurt, T Barrioz, M Barthet, C Boustiere, P Burtin, A Calazel, JM Canard, C Cellier, B Croguennec, PA Dalbies, R Laugier, JC Letard, B Marchetti, P Pienkowski, F Prat, B Richard-Molard, JC Saurin of the SFED

Introduction

An acute gastrointestinal haemorrhage is considered to be low gastrointestinal haemorrhage (LGIH) when it originates lower after the dudodeno-jejunal flexure. LGIHs represent less than 20 % of blood losses of the digestive system. In 95 % of cases, the cause is located to the sigmoid or rectum especially diverticular disease and vascular dysplasia, or from anal origine.

In all cases, hospitalisation is necessary except for minor hemorrhage due to anal wipping. The point is to rapidly detect a severe form which represents 10 % of cases and which may immediately impair the ultimate prognosis. The prognosis of digestive bleeding is essentially related to age and associated comorbidities. More than 80 % of LGIHs will spontaneously stop, but the recurrence rate remains high and may reach 15 to 20 % in case of diverticular disease.

The aetiological diagnosis of LGIHs remains a major issue. This search should start at an early stage as soon as haemodynamic stability is obtained.

It is important to distinguish lower rectal haemorrhage of proctologic origin from haemorrhage of upper origin with LGIH clinical presentation, especially when there exists haemodynamic changes, by performing an upper digestive endoscopy.

A lower rectal haemorrhage of proctologic origin can be suspected in the following conditions:

Bleeding associated with pain is the second reason for consultation in proctology. Bleeding, of anal origin, is never abundant and has no signs of severity, except if secondary to coagulation anomalies and/or of iatrogenic origin. Inquiry is crucial, specifying the existence of associated functional signs (anal pain at defecation, prolapse, rectal syndrome, etc.), trauma (thermometer, enema, sexual practices, etc.), previous local treatment and long-standing bleeding. Repeated low amounts of bleeding may induce a chronic anaemia.

Epidemiology

Epidemiological studies on LGIHs are rare. A North American study estimated the annual incidence for adults at around 20.5 for 100 000 inhabitants. In 2007, a prospective epidemiological study was carried out in France by the Association Nationale des Gastroentérologues des Hôpitaux Généraux (ANGH) in 1333 patients. The average age was 72 years ± 16 years, the sex ratio 1/1 and the ASA score 2.5 ± 0.9 with 50 % of ASA 3; the initial symptoms were proctorrhagia (red blood or wine colour in 93 % of cases or only purple in 7 % of cases). In 14 % of cases associated with a malaise, a drop of the systolic blood pressure (SBP) < 100 mmHg or a loss of consciousness were recorded. In 75 % of cases a predisposing drug intake (aspirin and/or platelet aggregation inhibitors in mono- or bitherapy, 33.5 %; anti-vitamin K, 21.2 %; non-steroidal anti-inflammatory drugs, 10.9 %; heparin or low-molecular-weight heparin, 7 %) was recorded.

Evaluation of severity

Evaluation of severity relies upon objective clinical signs of hypovolemia (orthostatic hypotension, tachycardia), signs of collapse and biological signs: hematocrit (Ht), better than using a normal complete blood count; a 3 points drop of the Ht corresponds to one blood unit. hypovolemic shock is patent only when blood losses represent more than 35 % of the total circulating blood mass. The severity and activity of bleeding can be easily assessed by the volume of blood units administered during the first 24 hto maintain a correct haemodynamic state (cut off: 6 blood units per 24 h).

Initial treatment

It is crucial to restore an efficient volemia through transfusion of blood units after set up of a venous access, in both arms, if needed. oxygen nasal supplementation may also be associated. Usually, an Ht > 25 % or a SAP > 80 mmHg are major objectives. Reaching a Hb rate > 7 or 10 g/100 ml depends on age and cardiorespiratory comorbidities.

Initial and minimal proctologic examination to rule out frequent anorectal causes

At the time bleeding remains well tolerated, proctologic examination should include anuscopy in a genu-pectoral or left lateral decubitus position without prior preparation and recto-sigmoidoscopy. This initial proctologic examination enables the viewing of the anal canal and lower rectal mucosa.

Only the observation of an active bleeding or an adhering clot has any formal diagnostic value at the time of examination. Additional subsequent work-up will be performed according to diathesis, abundance of bleeding and case history.

Aetiologic diagnosis A. Role of endoscopy In case of serious hematochezia, and associated with malaise or collapse, it is mandatory to perform first-line upper endoscopy, since anal red blood may also be due to a dramatic digestive hemorrhage from the upper digestive tract or to an accelerated time of the intestinal transit. Nasogastric intubation does not bring any additional contribution to exclude upper gastrointestinal origin. Despite in such cases a rent randomized trail comparing urgent versus elective colonoscopy was published, the result shows that upper endoscopy initially could rule out an upper GI source but then, use of urgent (> 12 h) colonoscopy showed no evidence of improving clinical outcomes or lowering cost as compared with routine elective (36 – 60 h) after presentation. In other cases, a complete proctologic examination will search an anorectal disorder, and a rectosigmoidoscopy, will be performed during the same session in a non fasting patient, without any oral preparation; preparation of the lower pathway through enema is discussed and depends on the objectives: discard an anorectal cause (no enema) or perform a flexible recto-sigmoidoscopy (possible preparation through enema particularly in the presence of abundant clots). Its importance is limited by its poor tolerance and low effectiveness, around 10 %. Moreover, a rectal or a sigmoid lesion does not exclude a lesion in the right colon and should systematically lead to a colonoscopy with ileoscopy at least in the last loop after standard colonic preparation (3 to 4 litres of PEG over 3 to 6 h), as quickly as possible. Despite, this examination may be performed urgently if the haemorrhage is abundant or recurring so as to obtain a endoscopic haemostatic procedure (clips, haemostatic injections, etc.). However, it requires a colonic preparation at least with colonic enema possibly with a water pump, but preparation via the upper pathway is to be favoured each time this is possible. Endoscopic signs for colonic origin haemorrhage are proposed: active haemorrhage, visible non-haemorrhagic vessel, adhering clot, fresh blood in a localised segment of the colon or rectum, diverticular ulceration with fresh blood in the proximity and absence of fresh blood in the ileum with fresh blood in the colon. The presence of a visible lesion (cancer, ischemic colitis, IBD) may be relevant as a cause of serious hematochezia, but rarely associated to, an active ongoing bleeding. Other diagnostic methods are only proposed when the colonoscopy has failed or is impossible and if the haemorrhage persists with a haemodynamic changes: In case the latter is absent, it is better to perform a repeat colonoscopy in the best conditions (total anaesthesia after and PEG preparation). B. Role of modern imaging: helical CT angiography and coelio-mesenteric arteriography Helical CT angiography within the first 24 h enables locating the bleeding in the colon or the small intestine in 50 % of cases and establishing the cause of the haemorrhage in 30 % of cases. It may help in the selection of patients for a contributive digestive angiography and lead, without delay, in the performance of a selective arterial embolisation. C. Role of technetium-marked red blood cell scan This technique is limited in such urgent situations and very poor on a diagnostic point of view. D. Investigation of the small intestine Upper and lower small bowel endoscopy do not have their place in a situation of emergency. In 5 – 10 % of cases, the aetiologic diagnosis remains uncertain. Following negative upper and lower endoscopy, recent studies have shown the usefulness of early examination of the small intestine with a video capsule after oral preparation with 2 L of PEG.

Aetiologic treatment

There is now an abundance of publications on endoscopic haemostasis of LGIH. They mainly rely on haemostatic clips, haemostatic injections and thermal methods ([Table 1]). The prioritisation of tasks in the management of the case often influences therapeutic modalities ([Fig. 1]).

Fig. 1 Decision tree for lower digestive haemorrhages. Table 1  Lesions Frequency Topography of lesions Evolution Lesion aspect Treatment Evaginations 50 – 60 % 80 % to G, but > 50 % to D bleeding Relapse 10 % to 2 years25 % to 5 years Diverticular neck– Visible vessel – Adhering clot– Active bleeding InjectionAPCClipsColectomy Angiodysplasia 5 – 10 % Caecum +++ Relapse 25 % Deep red stains, often multiple, 2 to 10 mm round, stellar, in the form of an arch APCInjection Iatrogenic Rectal ulce 5 – 23 % Anterior side of rectumAnoeactal junction Linear or punctiform InjectionAPCTrans-anal surgery (Point in X) Intestinal tumors 11 % Ischemic colitis 3 – 9 % Sigmoid, splenic angle Edemised, violet coloured and necrotic Symptomatic Inflammatory colitis 6 – 30 % Radio-induced rectocolitis 1 – 5 % Rectum anterior sideOne third inferior of the rectum Recurrent Red stains often multiple, round, stellar and fragile to contact APC Colonic Polypectomy 0.3 – 2.4 % Seat of the initial gesture As a jet or layer or ooze after detachment of adhering clot InjectionAPCClipsArterial embolisationSurgery Hemorrhoids 10 – 28 % Ligature Colonic or rectal varix Rare Ligature?Perineal membrane? Small intestine Vascular dysplasia, Meckel, diverticulum APC per enteroscopy APC: Argon Plasma Coagulation.

Conclusion

In spite of their frequent severity or emergency presentation, LGIHs have a favourable spontaneous outcome in 80 % of cases. Up to 15 % of hematochezia especially serious one, come form the upper digestive GI. Their severity is mainly associated with comorbidities and advanced age. Initial management relies on usual resuscitation measures so as to provide the best conditions to perform an aetiologic evaluation. Colonoscopy is the key examination procedure that enables diagnosis and a possible haemostatic treatment performed after upper endoscopy in serious cases. The major issues remain delay, bowel preparation and repetition if initial diagnosis and preparation has not been obtained. Then, the choice for a second- or third-line diagnostic ptocedure depends on local facilities, but aims at particularly identifying the colonic segment where the bleeding originated.

References

1 Naveau S, Franco D. Hémorragies digestives. In: Rambaud JC, ed. Traité de Gastroentérologie. Médecine Sciences Flammarion, 2000: 217 – 230

2 Facon A, Goldstein S. Prise en charge initiale des hémorragies digestives. In: Monographies de la Société Francophone de médecine d’urgence. ed. Pateron D, prise en charge des hémorragies digestives. Masson, Paris 2002: 21 – 33

3 Debette-Gratien M. Endoscopie digestive. In: Monographies de la Société francophone de médecine d’urgence. ed Pateron D, prise en charge des hémorragies digestives. Masson, Paris 2002: 36 – 56

4 Lesur G. Hémorragies digestives basses. Cahier de FMC. Gastroenterol Clin Biol 2003; 27: 1127 – 1126

5 Sautereau D, Pichon M. Prise en charge des hémorragies digestives basses abondantes. Post U 2004; 139 – 8

6 Pilleul F, Buc E, Dupas JL, Boyer J et al. Prise en charge de l’hémorragie d’origine diverticulaire. Gastroenterol Clin Biol 2007; 31: 8 – 9-C2: 47 – 52

7 Lesur G, Grigoriu I. Hémorragies digestives basses. EMC Gastroentérologie 2006 (9 – 006-A-11).

8 Bour B, Pillette C, Lesgourgues B et al. Hémorragies digestives basses aigues: résultats préliminaires d’une étude ANGH sur plus de 100 patients [abstract]. Gastroenterol Clin Biol 2008

9 Lesur G, Bour B. Hémorragies digestives basses. Acta Endosc 2009; 39: 394 – 396

10 Laine L, Shah A Am J Gast 2010; 105: 2636 – 2641